The kind of work I’m doing right now — my critical edition of Auden’s book The Shield of Achilles — is somewhat unusual, but some readers might be interested in the tools I’m using to get it done.

The first thing I did was to go to AbeBooks and order four copies of early editions of the book, two of the American edition (Random House) and two of the British (Faber). These need to be scrupulously compared for differences.

I selected one of them — the earliest, which means an American edition (the book came out here several months before it did in the U.K.) — and made it my working copy. Before annotating it, I took photos of every page of the book. Then I went through the book with a highlighter, marking every word or phrase that I believe will require annotation.

I grabbed a pencil and, on the pages and on sticky notes, made initial comments on ideas that need to go into my Introduction, calling attention to related passages.

Then I returned to the photos of the text. I opened the Photos app on my Mac, navigated to the photo of the first page, and typed the keyboard shortcut I use to invoke TextSniper. TextSniper is a fabulous app. When you invoke it you get an area-selection tool. Draw a rectangle around any text on your screen and TextSniper OCRs the text and copies it to your clipboard. There are other ways I could do this: for instance, I could scan the book into a PDF and then use an app like PDFpen to OCR the whole text. But that brings in a lot of extraneous material, for instance anything in the pages’ headers and footers. With TextSniper I get precisely the text I want — and it is the most accurate OCR tool I have ever used, by a long shot. So Photos to TextSniper to BBEdit — and very shortly I had a complete text of the book to work from.

Next: Markup — in Markdown. In this case basically headings and italics — pretty simple work that only took a few minutes. I went from a bunch of digital photos to a clean, accurate working text in little more than half an hour.

As soon as you start the work of textual editing you need to generate comments (about formatting, for instance) and queries for the eventual copy editor. And since Microsoft Word is the lingua franca of publishing, I therefore had to convert my Markdown file to Word. Most of the time I use pandoc for such conversions, but I find that Brett Terpstra’s Marked does a better job of preserving line breaks — and a book of poems has a lot of line breaks.

(So why not just paste the OCR’d text directly into Word, instead of using a text file as the intermediate stage? Because, as you surely know, structuring text in Word is a nightmare. You try to turn one line into a header and Word decides to make the next paragraph part of the header and change the typeface of the previous paragraph. And then you can’t figure out how to fix it. A plain-text file structured with Markdown is precise. My primary governing rule of writing and text-editing: Never open Word until you absolutely have to.)

Okay, so then I had my accurate, ready-to-be-annotated text in a Word file. Which left me with one final workflow problem to solve: adding the annotations, which in the published edition will appear at the end of the text. There are several ways to do this, involving split screens or external monitors or even second computers. But here’s what I did: I got out my little-used iPad and connected it to my MacBook Air with Sidecar. Now I can look at the Word file of the book’s text on the iPad and add annotations in BBEdit on the Mac. Baby, I got a stew going!